Paul-Studebaker

Self-diagnostics may continue to rely on crowd-sourced intelligence

April 16, 2019
Taking advantage of the diagnostics will depend on how well they’re engineered, and the experience and collaboration of the technicians who have to interpret them.

Our first car with a “check engine” light and a way to interpret it is a 1986 Honda Accord LXi. The top-line LXi came with fuel injection and a system to let the driver know something was amiss (the check engine light) and give the mechanic a hint what it might be. Find the indicator lamp on the electronic control unit (ECU) under the driver’s seat, count the number of flashes (1 to 13) between two-second pauses, and follow the diagnostic tree in the shop manual. It’s always the exhaust gas recirculation (EGR) system, and some fuel system cleaner clears it up for a while.

In 2005, we got our first car with second-generation onboard diagnostics—OBDII—so I bought a little $39 code reader that would tell me, for example, P1456. The factory shop manual would interpret that as something like “vacuum leak in evaporative recovery system,” and tell me to start digging through and testing hoses and solenoid valves. But plugging “Acura Integra code P1456” into Google immediately reveals the most likely cause as a loose or defective gas cap.

Around the time Honda built the Accord, BMW launched a new 8 Series coupe, equipped for the first time with CAN bus. The 8 Series is nosebleed expensive to both buy and maintain, so BMW uses it to try out new technology. Old 8 Series are rare and beautiful, but they’re a lot cheaper than they would be if they weren’t so complicated.

CAN bus lets BMW (and now, many other makers) do amazing things at reasonable cost. Optional accessories such as power seats (with heaters, massagers, coolers, etc.) have their own distributed “brains”—they just need power, ground and the bus connection, and to be configured in software according to the build sheet.

That intelligence includes self-diagnostics, so with the right software and connection, a technician can review the status of all the car’s systems, get detailed diagnostics, modify configurations and reset fault codes. I hear tell the average Hyundai now has much of this functionality accessible from the dashboard touchscreen.

BMW debugged and developed new technologies on the 8 Series and a few years later, put them in the less expensive, more common 7 Series, then the bread-and-butter 5 and 3 Series. We have a couple of 20-year-old 5 Series with about 80,000 and 160,000 miles on them. I bought them cheap because, while the mechanical and electrical systems are quite durable and reliable, they’re complex, and lots of little things go wrong. These set off one indicator lamp or another, driving the owners to newer models.

To stay on top of them, I have a basic, $250 Autel diagnostic tool that does a lot more than the OBDII code reader, but way less than a $2,500 professional tool. When a DSC light recently came on, I read the relevant code as “58 DSC precharging pump.” However, a quick Google search showed that code is most often caused by a deteriorated connection in the circuit board of the ABS module, not a problem with the $500 electromechanical pump. I’ve seen similar codes that turn out to be a faulty module relay, corroded connector or, in many cases, just a controller brain fart that resets and never returns.

On the BMW, the electronics in the ABS module are buried under the solenoid coils for the valves in the brake circuits. A new module is about $600, and has to be coded to the car serial number so it knows it’s controlling a particular model's engine, transmission and brakes. My Autel can’t do that, so add $300 and a trip to the dealer. That’s a lot, and a cottage industry has risen up that for $100-300, will “rebuild” the electronics in my module, test it and return it with the configuration unmolested, so I can just plug it in.

I expect similar things will become more common as the electronics in virtually everything in our lives and plants age and become less reliable. Taking advantage of the diagnostics will depend on how well they’re engineered, but also on the experience and collaboration of the technicians who have to interpret them.

About the author: Paul Studebaker
Paul Studebaker, Editor Emeritus, Control. Reitred from full-time employment in January 2020, Studebaker earned a master's degree in metallurgical engineering and gathered 12 years experience in manufacturing before becoming an award-winning writer and editor for publications including Control and Plant Services